When designing your IP recording solution there are many factors to consider when choosing your hardware to ensure you create a well-functioning system. It is crucial to examine your main objectives for the system. For example, do you want to view your video footage in real time or are you happy to playback the footage when an incident occurs? Or do you want the system to trigger an alarm in case of an intruder? Consider components such as the processor, storage and network as these will all impact on the performance of the system.
The processor can affect the video quality as much as the cameras. When choosing the NVR you need to make sure that the CPU can handle the required number of cameras and the resolution at which they record.
Here are some considerations:
- CPU load is higher during video coding and when motion detection tool is active.
- The higher the resolution and the frame rate the greater the CPU load.
- The more changes in the scene the camera detects the greater the load on the system.
A high end Intel® Core™ i7-4790 can handle 400fps@720p which is 33 1 mp cameras or 16 2 mp cameras at 12 fps. This is heavily dependent on the camera manufacture, camera performance, network, motion and environment (lighting, etc).
When we talk about storage for video surveillance most people think about this in terms of recording capacity. This is of course relevant when building a video security system, but the storage can have an impact on the performance as well; related to the number of disks used and how data are written on the disk.
Windows only allows a certain number of hard disks to be connected to it until it finds it hard to cope. When a system uses too many disks it can slow the performance of the entire server. We normally add a hard disk controller or RAID controller so there is no additional load on the CPU and this allows more hard disks to be connected.
Keep in mind that once the hard disk is formatted by Windows it will lose 7% of the available space so we would recommend you add an extra 10% to the amount of the space being used. The VMS software reserves by default 90% of available disk space on other drives than the C-drive for storage use. It does not use the C-drive for storage, except when it is the only drive. A separate disk for the operating system and application software is also recommended.
Another aspect when designing a good video recording system is the network. For Mirasys VMS, a Gigabit (1 Gbit/s) Ethernet is required on both the client and the server side. Data can be transmitted over the Internet, or any other network using TCP/IP.
The bandwidth requirements for the network between the recorders and client programs depend on different factors such as, the number of cameras, the resolution used, the video compression, frame rates and scene complexity. When designing your network, there are some considerations:
- Data transmission from the recorders to the client computers (uplink) requires more bandwidth than data transmission from the client computers to the recorder (downlink). The connection does not have to be symmetric.
- Real-time monitoring requires more bandwidth than transmitting recorded video or audio.
- In addition, signalling and protocols increase bandwidth usage.
Best practice with the system for security and network performance would be to have the cameras separated from the rest of the network. This can be done in two ways: having them physically separated, using camera local network switch that is connected to a recording server’s network adapter, or having them logically separated, using VLANs (Virtual LAN) on the switch. It is recommended that you consult an ICT expert or your IT department before planning and configuring VLANs on your VMS network’s networking devices. It is strongly recommended to have at least two 1 Gbit/s network adapters for each server: at least one adapter for the camera feeds and one adapter for server-client and server-server communications.
Bandwidth usage example:
Another best practise is to plan the physical network and maintain current topologies for record keeping. A clear network topology is useful in narrowing down network bottlenecks, failure points and sections that could be improved or expanded. It should include the cameras, recording servers, decoders, connection types, network switches, workstations, firewall or gateway servers, switches and routers. Mapping cabling and devices through a building or site can save a lot of headaches and time. The appropriate length of cables should also be considered to reduce interference or overheating.